Detection of early damage on strawberries using hyperspectral reflectance imaging combined with band ratio and improved watershed segmentation algorithm
In order to solve the problem of early damage of ripe strawberries during picking and transport, which is difficult to detect quickly, and to safeguard the commercial value and food safety of strawberries, this study proposes a new method based on hyperspectral imaging technology, which combines the band ratio and the improved watershed segmentation algorithm, to realize non-destructive detection of early damage of strawberries. Principal component cluster analysis was conducted across three spectral ranges: Vis-NIR (480.19~1 015.63 nm), Vis (480.19~780.0 nm), and NIR (780.0~1 015.63 nm). Characteristic wavelengths were identified using principal component loading maps. Subsequently, corresponding principal component and band ratio images were generated. The segmentation of strawberry damage regions was performed using an improved watershed segmentation algorithm. The results indicate that: 1) Strawberries with different damage areas can be effectively differentiated in the Vis-NIR and NIR spectral ranges, along the PC1 direction. 2) Among the band ratio images, the Q758.64/1 015.63 band ratio image exhibits the best performance in enhancing damage contrast, making it the most suitable for subsequent analysis. 3) The improved watershed segmentation algorithm outperforms the traditional watershed and improved Otsu methods, achieving an overall detection accuracy of 95.5%, demonstrating its reliability and effectiveness in detecting early strawberry damage. In conclusion, the integration of hyperspectral imaging technology with the improved watershed segmentation algorithm is a viable approach for the rapid, nondestructive detection of early strawberry damage, offering theoretical and technical support for ensuring strawberry quality and food safety.
LiuY, ZhouX Q, YuX F, LiY Q, HanS L. Research progress of nondestructive testing techniques for fruit and vegetable quality[J]. Journal of Zhejiang University :Agriculture and Life Sciences, 2020, 46(1):27-37 (in Chinese)
LiL, PengY K, LiY Y. Design and experiment on grading system for online non-destructive detection of internal and external quality of apple[J]. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(9):267-275 (in Chinese)
GengH, CaiC, LiuB. Prediction of apple internal quality based on dielectric properties[J]. Journal of Agricultural Mechanization Research, 2019, 41(10):186-191 (in Chinese)
ZuoJ W, PengY K, LiY Y, ZouW L, ZhaoX L, SunC. Study on detection and grading system of watermelon sugar content based on acoustic characteristics[J]. Transactions of the Chinese Society for Agricultural Machinery, 2022, 53(S1):316-323 (in Chinese)
[15]
BizzaniM, William Menezes FloresD, Bueno MoraesT, Alberto ColnagoL, David FerreiraM, Helena Fillet SpotoM. Non-invasive detection of internal flesh breakdown in intact palmer mangoes using time-domain nuclear magnetic resonance relaxometry [J]. Microchemical Journal, 2020, 158: 105208
XuS, LuH Z, WangY J, ZhouZ Y, JiangY M, LüE L. Storage time identification of kiwi fruit based on electronic nose and physical characteristic combination[J]. Food Science and Technology, 2016, 41(3):292-297 (in Chinese)
WangF, MengX Y, ChenL Y, DuanD D, QianY J. Research on apple damage detection based on hyperspectral imaging[J]. Guangdong Agricultural Sciences, 2023, 50(7):57-63 (in Chinese)
LiB, HanZ Y, WangQ, SunZ X, LiuY D. Research on bruise level detection of loquat based on hyperspectral imaging technology[J]. Spectroscopy and Spectral Analysis, 2023, 43(6):1792-1799 (in Chinese)
DuanJ L, ZhangH Y, FuH, LaiL Q, YuanH T, HeJ Y, YangZ. Collision damage detection of green bananas using hyperspectral imaging[J]. Transactions of the Chinese Society of Agricultural Engineering, 2023, 39(7):176-184 (in Chinese)
[30]
LiuQ, SunK, PengJ, XingM K, PanL Q, TuK. Identification of bruise and fungi contamination in strawberries using hyperspectral imaging technology and multivariate analysis[J]. Food Analytical Methods, 2018, 11(5): 1518-1527
[31]
FuX P, WangM Y. Detection of early bruises on pears using fluorescence hyperspectral imaging technique[J]. Food Analytical Methods, 2022, 15(1): 115-123
LiuQ, ZhangY Y, HuangJ, ZhaoL M. Applications and prospects of fruit image processing technology[J]. Journal of Qinghai Normal University, 2022, 38(4):36-44 (in Chinese)
[34]
YinH, LiB, LiuY D, ZhangF, SuC T, Ou-YangA G. Detection of early bruises on loquat using hyperspectral imaging technology coupled with band ratio and improved Otsu method[J]. Spectrochimica Acta Part A, Molecular and Biomolecular Spectroscopy, 2022, 283: 121775
OuyangA G, LiuH C, ChengL, JiangX G, LiX, HuX. Hyperspectral image features combined with spectral features used to classify the bruising time of peach[J]. Spectroscopy and Spectral Analysis, 2021, 41(8):2598-2603 (in Chinese)
WuW B, JinW B. Dimension reduction of hyperspectral images based on principal component analysis[J]. Inner Mongolia Science Technology & Economy, 2018(22):66-67 (in Chinese)
[39]
TianX, ZhangC, LiJ B, FanS X, YangY, HuangW Q. Detection of early decay on citrus using LW-NIR hyperspectral reflectance imaging coupled with two-band ratio and improved watershed segmentation algorithm [J]. Food Chemistry, 2021, 360: 130077
[40]
LiJ B, LuoW, WangZ L, FanS X. Early detection of decay on apples using hyperspectral reflectance imaging combining both principal component analysis and improved watershed segmentation method[J]. Postharvest Biology and Technology, 2019, 149: 235-246
LuN, HanP, WangJ H. Prediction on firmness of strawberry based on hyperspectral imaging[J]. Software Guide2018, 17(3):180-182 (in Chinese)
[43]
TianX, ZhangC, LiJ B, FanS X, YangY, HuangW Q. Detection of early decay on citrus using LW-NIR hyperspectral reflectance imaging coupled with two-band ratio and improved watershed segmentation algorithm [J]. Food Chemistry, 2021, 360: 130077
ZhaoM F, LiuZ D, ZouX, WuL, ZhangF, LongJ. Detection of defects on potatoes by hyperspectral imaging technology[J]. Laser Journal, 2016, 37(3):20-24 (in Chinese)
[50]
LiJ B, LuoW, HanL H, CaiZ L, GuoZ M. Two-wavelength image detection of early decayed oranges by coupling spectral classification with image processing[J]. Journal of Food Composition and Analysis, 2022, 111: 104642